Mechanisms Underlying the Use of Power-Creativity Relationship in the Military: Achievement Motivation and Identification

Abstract

ABSTRACTThese two studies tested how supervisors’ use of coercive or reward power influences individuals’ creativity and examined whether subordinates’ need for achievement affects this process. Dyad data were collected from two studies at two time points. The studies indicated that supervisors’ use of coercive or reward power affected subordinates’ identification with supervisor and thereby enhanced subordinate creativity. Interestingly, these results suggest that the functional effects of using coercion are positive if punishments are contingent upon performance. The pattern of the results showed that fear of failure moderated the relationship between supervisors’ use of coercive power and subordinates’ identification with supervisor. Subordinates’ hope for success strengthened the positive effects of supervisors’ use of contingent rewards on subordinates’ identification with supervisor, while these positive effects would be attenuated if supervisors used noncontingent rewards. Theoretical and practical implications of the study findings are discussed. Disclosure statementNo potential conflict of interest was reported by the author(s).Data availability statementThe data used in the study will be made available by the authors upon reasonable request.Notes1 We assessed the size of mediated effects by adopting Fairchild et al. (Citation2009) approach, which calculated the effect size of the mediation effect by dividing the indirect effect by the total effect. For Study 1, the effect sizes of (1) the mediation (i.e., Supervisors’ use of coercive power – Subordinates’ identification with supervisor – Subordinate creativity) is about 17% of this association and (2) the mediation (Supervisors’ use of reward power – Subordinates’ identification with supervisor – Subordinate creativity) is about 17% of this association. In other words, subordinates’ identification with supervisor acted as a mediator in the pathway between supervisors’ use of coercive/reward power and subordinate creativity, and accounted for about 17% of this relation (that is the effect size of the mediation effect). For Study 2, the effect sizes of (1) the mediation (i.e., Supervisors’ use of noncontingent coercive power – Subordinates’ identification with supervisor – Subordinate creativity) is 43% of this association and (2) the mediation (Supervisors’ use of contingent reward power – Subordinates’ identification with supervisor – Subordinate creativity) is 62% of this association. The ΔR2 effect sizes of Study 2 is much larger than Study 1. This difference indicates that supervisors’ use of power can better explain their effects on the mediator and dependent variable when the power bases (rewards and coercion) are differentiated into contingent or noncontingent types. Also, this difference suggests that being treated fairly by supervisors conveys evaluative information that makes subordinates feel respected, valued, and self-enhanced (Wiesenfeld et al., Citation2007). Thus, compared with receiving unspecified rewards or coercion in Study 1, subordinates identify more with a contingent reward supervisor who can afford instrumental benefits. Therefore, engaging in creative performance in return for their identified supervisors is promising for the future.2 Compared to most of past organizational behavior/human resources research, the moderation effect sizes in these two studies seem acceptable. Aguinis et al.’s (Citation2005) reported an average corrected effect size (f2) in tests of moderation = 0.002 from their sample of 196 meta-analyses, which can be converted as 0.0019 for an R2 value. Kenny (Citation2018) suggested that “perhaps a more realistic standard for effect sizes might be 0.005, 0.01, and 0.025 for small, medium, and large, respectively” with an optimistic hope in mind to get these f2 effect sizes.Additional informationFundingThis work was partly supported by the Ministry of Science and Technology (R.O.C.) under Grant (number 109-2410-H-606-004).